- SeeedStudio Grove Vision AI Module V2 Arm Cortex-M55 & Ethos-U55 (Arduino, RPi)
- High-speed 33ms inference time
- Versatile with CSI, IIC, UART, SPI, and Type-C interfaces
- Supports Pi cameras & PDM microphone
- Easy AI model training on the SenseCraft platform
- Low-power, battery-friendly design
The SeeedStudio Grove Vision AI Module V2 Arm Cortex-M55 & Ethos-U55 (Arduino, RPi) is a sophisticated MCU-based smart vision module powered by the Himax WiseEye2 HX6538 processor. It boasts a dual-core Arm Cortex-M55 and an integrated Arm Ethos-U55 neural network component. The module incorporates Arm Helium technology, which is finely optimized for vector data processing, enabling a significant uplift in DSP and ML capabilities while maintaining low power consumption, making it ideal for battery-powered applications.
The Grove Vision AI V2 has been evaluated against other MCU-based vision AI boards and has demonstrated superior performance with an impressive inference time of just 33 milliseconds and a fast frame rate of 30 FPS, providing a smooth experience. Its efficient power consumption and user-friendly interface make it a premium choice for high-performance vision processing tasks.
This module is now compatible with all Pi cameras through a standard CSI interface and is not limited to vision applications. It also features an onboard PDM microphone for sound applications and includes an SD card slot for convenient storage of images, videos, and identification results.
The board offers a variety of interfaces, including IIC, UART, SPI, and Type-C, allowing for easy connection to popular development platforms such as Seeed Studio XIAO, Grove, Raspberry Pi, BeagleBoard, and ESP-based products. This facilitates further development and seamless integration with various interfaces and data through programming environments like Arduino, Micropython, CircuitPython, and PlatformIO, as well as connectivity to the cloud or dedicated servers such as Home Assistance.
The SenseCraft AI platform supports easy AI model training and deployment with no-code/low-code options, ensuring complete adaptability of the trained models to Seeed products. The platform allows for immediate visualization of identification results on the website, enabling prompt assessment of model performance.
The Grove Vision AI V2 is particularly suited for tinyML applications, where it simplifies the addition of vision AI to smart sensors. Users can deploy off-the-shelf or custom AI models with ease, connecting the device, selecting a model, and viewing identification results without hassle.
Please note that the Grove Vision AI V2 does not include a camera, and users may need to acquire a compatible camera module separately.
